I have read a number of diaries expressing grave discontent with the poor state of comment moderation on DKos. I offer here a proposal for improving the process so that it serves its true purpose: to hide uncivil comments, elevate the level of discussion, and get rid of trolls.
There's no question that we cannot rely upon one or two or three moderators; there's simply too much activity taking place for so few moderators to stay on top of. The alternative now under test is the empowerment of a large community of Trusted Users who are presumed to have the best interests of the community at heart. Sadly, this hasn't worked out as planned; the system operates poorly because there are enough people who abuse it. The guidelines for using TU status are not often heeded.
There are a variety of solutions to this problem. Here are two:
PLAN 1. "The accusation is as serious as the crime." This is an ancient principle of primitive judicial systems meant to insure that accusations do not fly too wildly. If X accuses Y of a crime, and X fails to present a good case, then X is subject to the same penalties that Y would have been subjected to had Y been convicted. There were different definitions of what constituted a "good case", but scheme was, in a rough and ready way, workable.
If we were to use it here on DKos, then the act of HRing anything subjects you to the risk of having it rebound upon you. We would add one more button to the response buttons. The first and second allow you to Recommend or Hide Rate any comment. The new third button would activate only after somebody has HRed the comment. Trusted Users are asked to use the button if they believe that somebody has HRed a comment unfairly. This flags the comment for review by a "meta moderator". The meta moderator will examine the comment and rate it into one of three categories:
(a) certainly deserving an HR, in which case those who HRed it get a bonus recommendation point.
(b) questionable; no clear judgement possible. No action is taken.
(c) Certainly NOT deserving of an HR, in which case the HRs are revoked and those who HRed it are given HRs themselves.
This decision is made without regard to the opinions of any users; if a comment gains many recommendations and a few HRs, that has no bearing on the decision of the Meta Moderator. Only the rules are considered.
The problem with this proposal is that people might hit that third button every time they get an HR, bogging the Meta Moderator down with a burdensome caseload. This problem could be reduced in magnitude by applying a penalty of one HR to anybody who flags an HR that is subsequently determined to be in category (a). In other words, if a comment is HRed, and you ask the Meta Moderator to look at it, AND that Meta Moderator determines that the comment in question is certainly deserving of an HR, then you get one HR assessed against you as a penalty for wasting the Meta Moderator's time.
Remember, no penalty is ever assessed against anybody if the Meta Moderator believes that a reasonable case can be made in either direction. Penalties are assessed only when the MM is quite certain of the decision.
PLAN 2: A College of Judicious Users. The basic idea here is to set up a group of people somewhere between a handful of moderators and a mass of plebs. You can only get Judicious User status by requesting it. Your request will be reviewed by an actual human being, who will examine your history, including your diaries and comments, in an effort to determine how fair-minded, mature, and judicious you are. If you meet the high standards required of a Judicious User, you are offered a JU contract. This is an agreement in which you promise to abide by the published guidelines for using HRs. You print the agreement, sign it with your real name, and email a scan of the agreement back to DKos. Your public anonymity will, of course, be preserved. Only JUs can issue HRs; their HRs cannot be counterbalanced with Recs. Two HRs from two different JUs are sufficient to hide a comment; the threshold for autobanning a troll would be lower than now, because we can be more confident of the results coming from JUs.